introduction to economic evaluations - denver, colorado · last year’s ski boots (cheaper) than...

Introduction to Economic Evaluations

Marcelo Coca Perraillon

University of ColoradoAnschutz Medical Campus

Cost-Effectiveness AnalysisHSMP 6609

2016

1 / 36

Outline

Agenda for today: The big picture - an overview of this class

What is economic evaluation?

Several examples

Types of economic evaluations

Cost analysisCost-effectiveness analysisCost-utility analysisCost-benefit analysis

Overview of key concepts and elements of economic evaluations

Methods: ICER, decision trees/Markov models

Logistics and syllabus

2 / 36

Economic Evaluations

What is an economic evaluation?

(Textbook) “The comparative analysis of alternative courses of action interms of both their costs and consequences.”

Alternative courses of action: Health interventions like screeningprograms, new ways of delivering care, new medications, or newimaging technology. Also, intensity (dosage, screening frequency)

Comparative analysis: Always a comparison – current way of doingsthings versus new ways; old drugs versus new drugs; new imagingtechnology versus old imaging technology. A valid alternative couldbe “doing nothing”

Costs: Measured in $

Consequences: Outcomes measured in natural units (casesdetected/averted), years of life gained, years of life gained combinedwith a measure of quality (e.g. QALY), or even $

(As with most definitions, not totally right. Will explain in a bit)

3 / 36

Economic Evaluations

Purpose

Resources are scarce; we need a way to decide how to allocate them

The ultimate goal is to inform decision makers about the efficientallocation of resources

Economic evaluations provide a way to systematically think aboutcost and consequences of alternatives

An attempt to assess if an intervention or technology provides “value”

We will talk more about how economic evaluations are used (or not)in the US and in other countries later in the class

4 / 36

Examples

5 / 36

Examples

Bedrosian et al. (2015)

Does the use of MRI for women with invasive lobular carcinomachange the cost of care?

MRI improves staging → fewer operative interventions to achieve amargin-negative → fewer re-excisions → lower costs

On the other hand, MRI → delays and additional tests plus mixedevidence

Compared costs in women who received MRI vs those who did not atMD Anderson (piggyback clinical trial; about 100 patients)

Conclusion: MRI increases costs (but not a lot)

6 / 36

Examples

British Medical Journal (BMJ)

7 / 36

Examples

Patel et al. (2004)

Evaluate the cost effectiveness of (informal) caregiver training byexamining health and social care costs, informal care costs, andquality adjusted life years in caregivers

About 300 caregivers were randomized into two groups: training andcontrol group

They compared costs and benefits

Conclusion: Caregiver training during rehabilitation of patientsreduced costs of care while improving overall quality of life incaregivers

8 / 36

Examples

9 / 36

Examples

Heijnsdijk et al. (2015)

Objective: Assess the cost effectiveness of prostate cancer screening(PSA)

Used data from European Randomized Study of Screening forProstrate Cancer (ERSPC)

Simulation of clinical story of 10,000 men who are screening startingat 55

Predicted the numbers of 1) prostate cancers diagnosed, 2)prostatecancer deaths averted, 3) life-years and 4) quality-adjusted life-years(QALY) gained, and cost-effectiveness of 68 screening strategies forprostate cancer

Conclusion: Prostate cancer screening can be cost-effective when itis limited to two or three screens between ages 55 to 59 years.Screening above age 63 years is less cost-effective because of loss ofQALYs because of overdiagnosis.”

10 / 36

Examples

My examples

Colorado Family Planning Initiative

Targeted to low-income women at Title X clinicsProvided free long-acting contraceptives (LARCs) (IUDs, implants)Supported by an anonymous private foundationResearch question: what are the cost savings from the perspective ofthe state of Colorado?

Blood screening for celiac disease and diabetes

Large scale screening in children ages 2-17. UC-BDC plans to screen inDenver (and offer follow-up)Opportunity to measure (some) costs and benefitsThe intervention will last a couple of years but the effects have to betaken into account over a lifetimeAs in the prostrate screening example, we need to use simulations

11 / 36

Background

Key elements of economic evaluations

1 Perspective or point of view of the analysis: Who is incurring thecosts and consequences?

Provider, health department, societal, insurance company (private orMedicare/Medicaid)

2 Time of analysis (aka time horizon): short term, long term? Wholelifespan?

3 Discounting: Time preference. A benefit/cost now is not the sameas the same benefit/cost 30 years later (inflation/opportunity cost)

4 Sensitivity analysis: Evaluates the effects of uncertainty in numbers

5 Methods: Incremental cost effectiveness ratio, decision trees, Markovmodels

12 / 36

Background

Prerequisites of EEs

This may sound obvious, but...

You want to make sure the alternative you want to compare worksand does no harm (efficacy/effectiveness)

A story about economic evaluation proposals (or why most EEs neverhappen)

13 / 36

Background

Challenges

The fundamental challenge of EEs is... lack of DATA

Data synthesis (e.g. meta-analysis) is many times a key component ofEEs

Rarely a single sources of data; EEs combine multiple sources of data

I’ll try to provide examples of data sources for the US

14 / 36

Types of Economic Evaluations

Types of economic evaluations

How benefits are measured (or not) distinguishes different types ofeconomic evaluations

1 Cost-analysis (CA)2 Cost-effectiveness analysis (CEA)3 Cost-utility analysis (CUA)4 Cost-benefit analysis (CBA)

15 / 36


Cost analysis

Objective: compare the cost of two or more alternatives

Consequences are not examined (partial evaluation)

But... if benefits are the same→ it’s like comparing costs and benefits

A common version of cost analysis: cost of illness study (COI): COImeasures the economic burden of a disease(s) and estimates themaximum amount that could potentially be saved if a disease were tobe eradicated

We will spend some time in cost analysis because it’s common to alltypes of economic evaluations

16 / 36


More on costs

What costs should be considered depends on the perspective of theanalysis and the time horizon

Health sector: drugs, equipment, hospital stays of the program andpossibly follow-up

Other sectors: Social services, follow-up personnel, home health

Patient and/or family: Travel time, caregiver’s lost income, cost ofinformal care...

Productivity changes: Interventions may affect ability to work nowbut could increase the ability to work in the future

We’ll talk about these issues for at least two weeks...

17 / 36


Cost-effectiveness Analysis (CEA)

Objective: compare the cost and benefits/consequences of two ormore alternatives

Benefits are measured in “natural units”

Effectiveness measure should be relevant/important

Example of benefits: cases averted, births averted, cases detected,years of life gained, blood pressure, days free of pain...

Can compare several alternatives with same benefit measure

18 / 36


Cost-effectiveness Analysis (CEA)

Results presented as (incremental) costs per unit of outcome

Incremental Cost Effectiveness Ratio (ICER)

ICER =C1 − C2

E1 − E2

Ci and Ei are the costs and effectiveness measure of alternative i

We will spend some time discussing and calculating ICERs

Limitations: a) There is no preference/value associated with themeasure of benefit. b) Can only compare alternatives in which the“natural unit” of benefit is the same.

19 / 36


Cost-utility Analysis (CUA)

Same objective: compare the cost and consequences of two or morealternatives

Benefits are measured in quality-adjusted life years (QALYs)

QALYs combine both gains in extra years of life (quantity) with thequality of those gains

Can compare more alternatives

20 / 36


Cost-utility Analysis (CUA)

We will talk a lot about how economists measure “quality”(preferences)

As before, ICER provides a summary of results but now Ei = QALYi

ICER =C1 − C2

QALY1 − QALY2

Limitation: not all measures of benefits are transformed intopreferences or quality

21 / 36


Using ICER

We have ICER comparing alternative B to usual care, A. Now what?

Easy cases:

1 B is more expensive and less effective (prefer A)2 B is less expensive and more effective (prefer B)

Not-so-easy cases:

1 B is more expensive and more effective (how do we know if extrabenefit is worth the extra cost?)

2 B is both less expensive and less effective (how do we know if the costsavings are worth the reduced effectiveness?)

22 / 36


The cost-effectiveness plane

23 / 36


Back to the harder cases

A) Is the extra benefit worth the extra costs or B) Are the costssavings worth the reduced benefits?

There is an implicit notion of value. Think about how you makedecisions in your life.

Is it worth it to get an MPH?Last year’s ski boots (cheaper) than this year’s model (moreexpensive)?

Key is what we are giving up and how we value it

The notion of opportunity cost: the loss of potential gain fromother alternatives when one alternative is chosen

The notion of a threshold value. The (in)famous $50,000 per QALY.

We will come back to these issues during the semester

24 / 36


Cost-benefit Analysis (CBA)

Once again, same objective: compare the cost and consequences oftwo or more alternatives

But now benefits are measured in monetary units

CBA is the type of economic evaluation that is more consistent witheconomic theory

CBA can tell us if an intervention is worth doing

The problem of using CBA in health is that it is very difficult totransform measures of benefits into monetary values

25 / 36


Cost-benefit Analysis (CBA)

What is the dollar value of a case averted? What is the value of anextra year of life? What is the value of not being able to walk?

This doesn’t mean that economists don’t try... We will study ways toassign monetary values to (some) benefits

CBA is extensively used in certain project evaluations (infrastructure)but not so often in health

If benefits > costs, then it is worth it to do the project

Summary measure: benefit/cost ratio or (benefit − cost)

26 / 36


A note on terminology (jargon)

“Economic evaluation” is the more general term. It may refer to oneof the four types of studies we discussed (CA, CEA, CUA, and CB)

“Cost-effectiveness analysis” is also used as a general term but itrefers to CEA and CUA. Blame your textbook (up to third edition)for insisting on separating CEA and CUA

There is always confusion with names – this class is called CostBenefit and Effectiveness in Health...

More important than definitions is understanding the principles andmaking the right decisions

Remember, EE is about making decision on how to allocate resources

27 / 36

Methods

Methods

We have talked about how to calculate ICER

But there are other ways of conducting economic evaluations:decision trees/Markov models

Think of them as alternative ways of calculating ICER, but ways thatallow us to incorporate uncertainty more directly

It’s going to be clearer by the end of the semester, don’t worry

28 / 36

Methods

Decision analysis

Defintion: A systematic approach to decision making underuncertainty (Raiffa 1968)

It originated as a way of figuring out how we should make decisionswhen there is uncertainty (that is somehow quantifiable)

Here is the confusing part: it can also be used to perform aneconomic evaluation study

Another definition: Decision analysis is a way to gather all theinformation needed to perform an economic evaluation, includinguncertainty about effects

Your textbook does not cover much on decision analysis; we willcomplement it with other readings

We will use Excel; TreeAge is the most popular software but it’s notthe best tool for learning (black box). Excel is actually more flexible(but requires more programming)

29 / 36

Methods

Decision Tree

30 / 36

Methods

Decision Tree

31 / 36

Methods

Markov Models

Useful when events are recurrent (like migraines)

Essentially, a recursive decision tree with cycles/periods

Mutually exclusive “states” representing possible consequences

Simulate a cohort going from one state to the other in each cyclewith certain probability

Each state may have a cost and/or benefit associated with it, whichtransforms Markov models into an economic evaluation

32 / 36

Methods

Example: Burlone et al. (2013)

Evaluating the effects of extending contraceptive coverage under the ACAin OR

33 / 36

Importance

Why this class should be useful

Although CEA is not often considered as a basis for making decisionsin the US, the field is advancing rapidly

Several studies have influenced practice (e.g. screening for cancer)

Techniques to allocate costs are helpful in the real world

Decision analysis tools are also helpful in the real world

We will talk about ways to measure health (the benefits part of CEA)

And there is of course the joy of learning

34 / 36

Objectives

At the end of the class you will be able to:

Critically assess an EE evaluation study

Replicate an EE study

Present results of an EE study

Know where to start if asked to do an EE (then call a friend)

The school of pharmacy offers a more advanced class PHSC 7611:Applied Cost-Effectiveness Analysis

35 / 36

Logistics

Textbook and homeworks

The great thing about the textbook is that it covers pretty much everyimportant topic in EE. The problem is that it covers every topic

The class notes are your guide to the key parts in each chapter butyou should read the assigned chapters

Read chapter before or after class?

We will have homework most weeks. First homework due next week.Choose articles wisely

Groups

Excel

Office hours (do stop by)

36 / 36

introduction to economic evaluations - denver, colorado · last year’s ski boots (cheaper) than...

Documents